AI is the renovation, not the asset

Do you own a service business that makes steady money, runs the same way it did ten years ago, and gets a phone call from a buyer you have never heard of about once a quarter?

The calls are not random. Private equity and venture money has found the boring end of the economy — accounting practices, trades, property managers, anything with a phone that rings all day and a back office full of manual work. A Forbes report in June said Thrive Holdings, spun out of Joshua Kushner's Thrive Capital, is committing $1 billion to buying local accounting practices, after picking up nearly fifty of them in two years.

Here is the uncomfortable part. The person on the other end of that call has already worked out what your business would earn with its back office fixed. You probably have not.

If you own one of these businesses, that is the position you are sitting in.

The money is in the asset, not the AI

Strip the headlines off an AI roll-up and it is the oldest play in property and hospitality. Buy an under-managed asset. Fix the operation. Lift the margin. Sell it, or hold it, at a better multiple than you paid.

The buyer pays for the business on today's margin. They sell it on tomorrow's. The spread between those two numbers is the whole return — and AI is simply the tool that closes it faster than a new manager and a better spreadsheet used to.

AI is the renovation. The asset is still the asset.

That gives you two honest positions to take, and both start from the same piece of arithmetic.

• Become the acquirer. You already know your sector better than any fund does. Look at the tired operators around you — the ones with loyal customers, an owner nearing the end, and a back office still running on paper and goodwill. You can see the fix from the car park. They can't.
• Get ready to be acquired. Work out what your margin would be if someone else fixed your back office, then fix enough of it yourself that the buyer can't pocket the whole difference. Otherwise you are selling them the renovation at the price of the ruin.

The honest objection is real. Roll-ups fail, and they usually fail in one of two ways.

The buyer overpays, so no amount of margin improvement ever earns the price back. Or the buyer strips out the people customers actually trusted — the partner who knew every client's family, the tradesman who always turned up — and the revenue walks out the door behind them.

Automate the paperwork and the business gets better. Automate the relationship and it gets smaller.

Constellation Software has done this for thirty years

Mark Leonard spent more than a decade in venture capital before founding Constellation Software in Toronto in 1995. The idea was unglamorous on purpose: buy small vertical-market software businesses that serve one narrow industry each, keep them decentralised, and hold them.

It became one of the great compounding stories in public markets. By the time Leonard stepped down as president in September 2025, citing his health, Constellation had bought hundreds of these businesses and was worth tens of billions of dollars.

What made it work was not the software. It was the hurdle rate. Leonard's shareholder letters kept coming back to the return every acquisition had to clear before the company would buy it, and the company held to it.

The Globe and Mail reported in 2024 that bankers had come to believe Constellation would never pay a premium, and often stopped inviting it into auctions at all. Leonard's own letter that year conceded the point — the high hurdle was costing them deals, and he lowered it for larger acquisitions.

It is tempting to call that a technology story. It isn't. Constellation lost deal after deal on purpose, because the price it paid was where the return was actually made.

The discipline was never in the product. It was in the number they refused to go past.

Why I do it this way

I do this, and I have for years. I just did it before anyone called it a roll-up.

Most of my career has been as an owner-operator in serviced apartments and a venue, with full profit and loss responsibility. That work is acquiring operating assets and repositioning them.

The value was rarely in the building. It was in how the building ran — the rostering, the costs nobody had questioned, the standards that had drifted. Fix those and the same asset is worth more without a single new brick.

The lesson that stuck was the one Constellation learned. The deals that went well were the ones where I paid a price the operation could actually earn back. And the people guests trusted stayed, because they were half of what I had bought.

Start here this week

Take last year's profit and loss. Mark every line that is back-office work — bookkeeping, scheduling, invoicing, chasing, answering the same questions on the phone — and write next to each one what it would cost if it were done properly with the tools available today.

Add up the difference and put it on one page with your current margin at the top and the fixed margin underneath.

By Friday you will know the number the buyer already knows. Then decide whether you want to be the one who closes that gap — in your business, or in somebody else's.